From mboxrd@z Thu Jan 1 00:00:00 1970 From: Thomas Monjalon Subject: Re: [PATCH v2 06/11] app/test: use linked list to store PCI drivers Date: Thu, 12 May 2016 18:08:16 +0200 Message-ID: <1659775.2ddTAuyrkI@xps13> References: <1462904011-29838-1-git-send-email-viktorin@rehivetech.com> <2242767.BO2sf06KAx@xps13> <20160512175307.0573297e@pcviktorin.fit.vutbr.cz> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7Bit Cc: dev@dpdk.org, David Marchand , Bruce Richardson To: Jan Viktorin Return-path: Received: from mail-wm0-f47.google.com (mail-wm0-f47.google.com [74.125.82.47]) by dpdk.org (Postfix) with ESMTP id A54376CAF for ; Thu, 12 May 2016 18:08:18 +0200 (CEST) Received: by mail-wm0-f47.google.com with SMTP id n129so265800707wmn.1 for ; Thu, 12 May 2016 09:08:18 -0700 (PDT) In-Reply-To: <20160512175307.0573297e@pcviktorin.fit.vutbr.cz> List-Id: patches and discussions about DPDK List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org Sender: "dev" 2016-05-12 17:53, Jan Viktorin: > On Thu, 12 May 2016 17:31:28 +0200 > Thomas Monjalon wrote: > > > 2016-05-10 20:13, Jan Viktorin: > > > The test unregisters all real drivers before starting into an array. This > > > inflexiable as we can use a linked list for this purpose. > > > > I don't understand this. Maybe some words are missing. > > Better? > > The test unregisters all real drivers before starting (stored into an array). > This is inflexiable (due to its fixed size) and we can use a linked list for > this purpose. Better with a past tense? "The test was unregistering..." > > > +/* real drivers (not used for testing) */ > > > > What do mean by "not used for testing"? > > The test now avoids the DPDK builtin drivers. It only considers its > internal fake drivers my_driver and my_driver2. So the real drivers > are temporarily store into the real_pci_driver_list and returned back > after the test finishes. Maybe adding "Save" or "Backup" would make it clear. > It is the linked list mentioned in the commit log. It replaces the > original fixed-size array. > > (For drivers, it does not matter that much. But for devices, I think, > it is not a good practice to consider them in autotests. Every PC > where you execute the tests have different set of PCI devices.) Yes